Well, I'm a carb guy, and since Google brings up only about 500,000 pages when "how the rotary engine works" is typed into the search field, I'm not going to even attempt to explain the whole thing. With so many complete efforts already existing on long standing web sites that serve as iconic examples of rotary education, it makes little sense to delve too deeply into redundancy.

I'm also not going to get into the myriad of different size & porting configurations, nor delve into boosted applications (at least not here on this page), and I'm not going to talk about rotary timing or lubrication. These are each very important parts of maintaining and optimizing the performance of the rotary engine that require their own dedicated pages of information, none of which I'm qualified to give. I'm going to concentrate only on my own specialty, carbureted induction, and leave the rest to the experts of those fields.

But I do happen to have a few nifty little graphics, and since a picture's said to be worth a thousand words, I'll just breeze through the "how it works" part and concentrate my efforts on the breathing requirements of the rotary engine. Besides, if you don't know how your special little engine works, you should definitely be visiting some sites that have much better explanations of the rotary than I'm qualified to give.

Here goes my super-short synopsis...

This illustration shows the exploded view of the main components of the Mazda rotary engine.

Here we see the very simplified comparison between the Otto 4 cycle engine and the Wankle rotary engine.

So for the purposes of understanding the intake requirements of the rotary engine, let's study the cycles of both the Otto and the Wankle. If you keep an eye on the main shafts of each engine in the animation, you should see that the Wankle rotary main shaft, or eccentric shaft, rotates one full revolution for every power stroke, whereas the Otto piston engine main shaft, or crank shaft, rotates twice for every power stroke. It can be a bit tricky gazing at the rotary illustration, but keep your eye on the shaft with regards to the spark plug.

The reason this is so important is that when we compare the breathing requirements for a 70 CID piston engine like that found in the old Ford Pinto, and the 70 CID 12a rotary engine found in early Mazda sedans and Rx-7s, we can see that the rotary needs twice the intake per revolution. Factor in the that the rotary redlines at 8400 RPM (according to Racing Beat), and you have need for an even larger carburetor.

Whenever I refer to "the rotary" throughout these pages in any manner of quantification, unless specifically stated otherwise, just assume I'm referring to the Mazda 12a rotary, since that comprises the bulk of the SMW Nikki carburetor applications. Usually I'll simply write "the 12a", assuming that most readers already know what that is. In any mathematics illustrated here with regards to the naturally aspirated Mazda 12a rotary engine, unless otherwise specified, take for granted that the pressure is atmospheric at sea level, which is 14.7 lbs per square inch.

Also, please keep in mind that these pages are written in an effort to explain rotary engine carburetion in as comprehensive a manner possible to anyone with limited or no knowledge of the needs of the rotary engine, the basics of carburetion, or basic physics. It is not my intention to assume that most readers do or do not have an understanding of any of those topics. However, in an effort to not take for granted that the reader most certainly does already have a background knowledge in any of these topics, I may tend to "over explain" or be redundant throughout these pages. My only intention is to present the reader with a comprehensive understanding of the dynamics of the carburetor with regards to the rotary engine.

One more thing; What follows on this page and several carburetor theory pages is math used to illustrate theory. It's important to remember when doing "theory math" that "real world math" has far many more variables contributing to it than the very simple, straight forward math used to illustrate the theory behind things like flow dynamics and capacities. Theoretical math is good for demonstrating the maximum limits of something, but certainly not good for use as a foundation to build any "real world math" upon. The fact is, unless extensive physical testing is actually done on a model, the theoretical numbers can only serve to show us the absolute limits of what can be done.

Now that that's out of the way, let's get started...

With regards to induction, the most important difference is how we calculate the intake requirements of these two engines. We see by the illustration that the Otto engine's crank shaft rotates twice for the full displacement intake, whereas the Wankle engine's eccentric shaft rotates only once. This is a fundamental difference when comparing the intake requirements of both engines of similar displacement, though another important difference is that the Wankle tends to have a higher maximum RPM than most Otto engines of similar displacement of the same era.

As your engine purrs at idle, the rotors sweep by, taking in all the air that can fill up the chamber. The chamber can only theoretically ever be filled up with as much air as it can hold, and is actually limited by the amount of air the intake port will allow into the chamber in the time it takes for the rotor to pass by. The engine does not suck air in, it merely creates a void for the air to occupy. Since it does take some time for the air to occupy the vacancy created by the rotor, as RPMs increase, volumetric efficiency (the percentage of the chamber that air can fill during the intake stroke) often decreases, particularly with Otto cycle (piston) engines. Only changing the intake and exhaust port shapes, sizes and timing can effect volumetric efficiency (henceforth referred to as "VE") in a naturally aspirated rotary engine.

The exhaust system of an engine needs to be mentioned here with regards to engine breathing capacity. If the exhaust system is not "free flowing", it will restrict the ability of the engine to pass air through it. Exhaust is hot gas and as such has far more volume than the incoming air. Thus it requires large ports and a large output system to allow the engine to expel that massive volume. If the engine is stifled by a restrictive exhaust, it won't matter what carburetor you install, it simply won't be able to take in it's full capacity. The pressure of the expanded exhaust gases will overwhelm the pressure created in the chambers and partial cavitation will occur. Basically, the engine will run out of power. The RPMs may still rise, but the mixture will be very lean and overheating becomes a concern.

The carburetor's just a big paperweight until there's air flowing through it. To get the air into the engine you don't need a carburetor, you need the carburetor to get the fuel into the engine. Liquid fuel of any kind, no matter how volatile, still needs to mix with oxygen in order to burn. It needs to be mixed with a very specific amount of oxygen in order to combust. While fuel readily mixes with oxygen on a low enough level to burn just by evaporation, it doesn't physically mix in the specific ratios needed for combustion. The air to fuel ratio required for gasoline to combust is such that there's too much fuel to chemically mix with the air, so it needs to be atomized, or vaporized. Because it doesn't stay that way very long, the vapor needs to be introduced into the engine quickly in order for combustion to occur. The useable air to fuel ratios for the rotary range from 10:1 to 18:1. (* The 10:1 ratio is specific to the idle on Rx-7 rotaries with original emissions controls in place. It is exceptionally rich due to the exhaust gas recirculation process necessary for the complete burning of gasoline byproducts.)

Now that we've established how air needs to flow through the engine as well as the fuel requirements for the rotary, it's time to take a good look at just how much air we can expect the rotary to ingest. Here is where we decide what size carburetor to select. A carburetor any larger than what is required will only result in an engine with less than optimal performance. The reasons are explained in the Carburetor Basics pages.

The 12a rotary is a 70 cubic inch engine. That's it, just 70 cid. The Ford Pinto has a four cylinder 70 cubic inch engine, but since it's an Otto cycle engine, it's only theoretically capable of combusting 35 cubic inches per rotation of it's main shaft. The rotary is theoretically capable of combusting 70 cubic inches per rotation. Let's do some math...

The usual calculation for determining the maximum air flow for a piston engine is thus:

 

( engine displacement ) x ( maximum RPM )	=   Maximum CFM
3456

In the US, engine displacement is measured in cubic inches. "RPM" refers to revolutions per minute, and the denominator converts the product of cubic inches per minute into cubic feet per minute and takes into account the fact that the Otto engine rotates its shaft twice per intake stroke. The denominator of "3456" is actually the 12 inches x 12 inches x 12 inches found in 1 cubic foot (1728), multiplied by a factor of two, which effectively halves the overall equation. This same equation can be written like this:

 

(½)	( engine displacement ) x ( maximum RPM )	=   Maximum CFM
	1728

This is how the equation pertains to the Wankle rotary:

 

( engine displacement ) x ( maximum RPM )	=   Maximum CFM
1728

-It's about as straight forward as you can get. The 70 cid Mazda 12a Wankle rotary has two rotors which displace 35 cubic inches each. In one eccentric shaft rotation the two rotors draw in a combined 70 cubic inches. This is why the product is divided directly by the number of inches in a cubic foot, and not further divided in half, as done with the Otto engine.

Now let's talk about something all too often overlooked while calculating the intake of an engine; Volumetric efficiency.

Air ( or any gas or fluid ) flowing through a vessel is going to have some friction that will decrease the efficiency of that flow. The more complex the vessel, the more friction. A vessel such as a short, large diameter tube is going to flow air easily with very little friction. Fit a small pipe to one end of it, and the flow efficiency is going to decrease dramatically. Do the same to the other end, and add a few moving parts inside, and well, now volumetric efficiency becomes a serious part of the equation that determines the flow capability of the vessel.

Volumetric efficiency, or "VE", is a measurement of how efficient the flow through an engine is, given as a percent. Obviously, 100% VE is basically perfect, though specially designed and ported engines can actually reach beyond the 100% mark. However, in modified production engines these occurrences are so rare and contingent upon specific circumstances that happen very briefly, they are usually dismissed in dyno data as anomalies that will only serve to corrupt overall data averages.

With regards to the stock port Wankle rotary with a unrestrictive intake and header exhaust, VE percentage is in the mid to high 80s. Porting the rotary engine does not increase displacement. Porting increases VE, and by increasing flow efficiency, the engine can make more power. Attaining an overall volumetric efficiency of better than 95% throughout most of the power band in a rotary is something only the best builders can accomplish. It's not something that happens by accident. But even the average "street porting" can yield a 5 - 7% increase in VE. For a 12a with 85% VE, a street porting job that brings the VE up to 90% is actually an increase of about 6% of what it was.

Let's add the VE variable to complete the equation for the rotary air flow:

 

(VE)	( engine displacement ) x ( maximum RPM )	=   Maximum CFM
	1728

Now we'll get down to business and have some fun plugging in real numbers to see what our rotary engines really need in the way of air flow. Let's use the 12a again, and say it has an average VE of 85%. We'll give it the maximum revolutions per minute that Racing Beat suggests we never go past in a stock engine, which is 8400 RPM.

 

(85%)	( 70 ci ) x ( 8400 RPM )	= 289 CFM
	1728

Yeah, that's right. Simply doing the math should be enough to make anyone think twice about slapping that giant 750 CFM Holley on their 12a. But I'll get into all that in the Carburetor Basics section. Right now let's plug in a little better VE and see what we get...

 

(90%)	( 70 ci ) x ( 8400 RPM )	=  306 CFM
	1728

What happens if we port the 12a so well that the VE is 100%, and we pin the stationary gears so the maximum RPM is up to 10,000?

 

(100%)	( 70 ci ) x ( 10,000 RPM )	=  405 CFM
	1728

Yeah, no kidding! Let's do the same with the 80 cid Mazda 13b rotary...

 

(100%)	( 80 ci ) x ( 10,000 RPM )	=  462 CFM
	1728

I wrote up top that there's a difference between "theory math" and "real world math". These equations are theory math, and though they illustrate the limits of the intake capabilities of the stock port 12a & 13b rotaries, they fail to take into account the flow dynamics of the exhaust, the intake manifold, or the carburetor. Each of these components by themselves are quite complex in their own right, and subtle changes in each or any combination of the three will directly effect the volumetric efficiency and power output of the engine.

There is one more very important part of the engine combustion that must be addressed as a significant variable for performance as it will need to be optimized according to components and application, and that's the ignition system. The Rotary Ignition page will be written by someone who knows far more than I do about it. It can be treated as an entirely separate issue, and we can go on to Carburetor Basics from here. But ignition timing is extremely important, and I have had customers disappointed with the performance of a new carburetor I've sent them because their ignition was set wrong.